Initial vestibulo-ocular reflex during transient angular and linear acceleration in human cerebellar dysfunction

During transient, high-acceleration rotation, performance of the normal ves tibule-ocular reflex (VOR) depends on viewing distance. With near targets, gain (eye velocity/head velocity) enhancement is manifest almost immediatel y after ocular rotation begins. Later in the response, VOR gain depends on both head rotation and translation; gain for near targets is decreased for rotation about axes anterior to the otoliths and augmented for rotation abo ut axes posterior to the otoliths. We sought to determine whether subjects with cerebellar dysfunction have impaired modification of the VOR with targ et distance. Eleven subjects of average age 48+/-16 years (mean +/- standar d deviation, SD) with cerebellar dysfunction underwent transients of direct ionally unpredictable whole-body yaw rotation to a peak angular acceleratio n of 1000 or 2800 degrees/s(2) while viewing a target either 15 cm or 500 c m distant. Immediately before onset of head rotation, the lights were extin guished and were relit only after the rotation was completed. The axis of h ead rotation was varied so that it was located 20 cm behind the eyes, 7 cm behind the eyes (centered between the otoliths), centered between the eyes, or 10 cm anterior to the eyes. Angular eye and head positions were measure d with magnetic search coils. The VOR in subjects with cerebellar dysfuncti on was compared with the response from 12 normal subjects of mean age 25+/- 4 years. In the period 35-45 ms after onset of 2800 degrees/s(2) head rotat ion, gain was independent of rotational axis. In this period, subjects with cerebellar dysfunction had a mean VOR gain of 0.5+0.2, significantly lower than the normal range of 1.0+/-0.2. During a later period, 125-135 ms afte r head rotation about an otolith-centered axis, subjects with cerebellar dy sfunction had a mean VOR gain of 0.67+/-0.46, significantly lower than the value of 1.06+/-0.14 in controls. Unlike normal subjects, those with cerebe llar dysfunction did not show modification of VOR gain with target distance in the early response and only one subject showed a correct effect of targ et distance in the later response. The effect of target distance was quanti tatively assessed by subtracting gain for a target 500 cm distant from gain for a target 15 cm distant. During the period 35-45 ms after the onset of 2800 degrees/s(2) head motion, only two subjects with cerebellar loss demon strated significant VOR gain enhancement with a near target, and both of th ese exhibited less than half of the mean enhancement for control subjects. During the later period 125-135 ms after the onset of head rotation, when V OR gain normally depended on both target location and otolith translation, only one subject with cerebellar dysfunction consistently demonstrated gain changes in the normal direction. These findings support a role for the cer ebellum in gain modulation of both the canal and otolith VOR in response to changes in distance. The short latency of gain modification suggests that the cerebellum may normally participate in target distance-related modulati on of direct VOR pathways in a manner similar to that found in plasticity i nduced by visual-vestibular mismatch.